摘要
以李子柒为代表的中国田园生活类自媒体短视频,在国际社交媒体平台上受到海外网友的追捧和喜爱,为新时代我国如何推动中国文化走出去提供了参考路径和示范经验。运用卷积神经网络(CNN)和潜在狄利克雷分布(LDA)主题模型对李子柒短视频下的评论文本进行文本挖掘和情感分类,可快速识别文本中的情感倾向,了解海外受众在观看李子柒的视频后对中国文化产生的文化情感和文化认知,从而为我国主流媒体如何利用好海外社交媒体平台和短视频,向世界讲好中国乡村故事、传播中国乡土文化提供借鉴。Represented by Li Ziqi, short videos about Chinese pastoral lifestyle have garnered praise and affection from overseas internet users on international social media platforms. They provide a reference and exemplary experience for how China can promote its culture abroad in the new era. By employing Convolutional Neural Networks (CNN) and Latent Dirichlet Allocation (LDA) topic models, text mining and sentiment classification are conducted on comments over Li Ziqi’s short videos. This enables rapid identification of emotional tendencies in the text, understanding overseas audiences’ cultural emotions and perceptions towards Chinese culture after watching Li Ziqi’s videos. Thus, it offers insights for mainstream Chinese media on effectively utilizing overseas social media platforms and short videos to tell compelling Chinese rural stories and promote rural Chinese culture to the world.
出处
《数据挖掘》
2024年第4期207-217,共11页
Hans Journal of Data Mining